Vercel adds Grok 4.7, cuts prices 40%
Vercel has integrated SpaceXAI's Grok 4.7 into AI Gateway, the fx CLI agent, and the eve framework, offering 40% off inference through September 27. The model features a 500K token context window with adjustable reasoning levels from low to xhigh, zero data retention, and no platform markup on BYOK calls.
Vercel's immediate day-one rollout and 40% subsidy for Grok 4.7 signal that routing gateways are aggressively weaponizing price promotions and native agent tooling to win developer mindshare over direct provider APIs.
- –Deep reasoning meets massive context: The 500K context window paired with granular reasoning levels from low to xhigh targets complex, multi-hour coding workflows where standard frontier models become cost-prohibitive.
- –Gateway ecosystem lock-in: Seamless integration across AI Gateway, the fx CLI, and the eve agent framework demonstrates Vercel's strategy to capture the entire agent orchestration stack rather than serving as a dumb proxy.
- –Price wars at the inference layer: A 40% discount paired with zero platform markup on BYOK requests creates strong economic pressure for teams to route production traffic through Vercel rather than SpaceXAI directly.
- –Enterprise safety out of the box: Native Zero Data Retention (ZDR) and guaranteed disallowal of prompt training remove enterprise compliance friction from testing SpaceXAI's latest architecture.
DISCOVERED
1h ago
2026-09-21
PUBLISHED
1h ago
2026-09-21
RELEVANCE
AUTHOR
vercel_dev